SS-MAC: A Novel Time Slot-Sharing MAC for Safety Messages ... · Messages (BSMs) [2] to all neighbors within one-hop. By being constantly aware of the events in their surrounding

0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2780829, IEEETransactions on Vehicular Technology

1

SS-MAC: A Novel Time Slot-Sharing MAC forSafety Messages Broadcasting in VANETs

Feng Lyu, Student Member, IEEE, Hongzi Zhu, Member, IEEE, Haibo Zhou, Member, IEEE, Wenchao Xu,Ning Zhang, Member, IEEE, Minglu Li, Senior Member, IEEE, and Xuemin (Sherman) Shen, Fellow, IEEE

Abstract—Efficient and scalable media access control (MAC)protocol design is crucial to guarantee the reliable broadcastof safety messages in vehicular ad hoc networks (VANETs). Todevise a MAC for safety message broadcasting with reliabilityand minimum delay, in this paper, we propose a novel time slot-sharing MAC, referred to as SS-MAC, which can support diverseperiodical broadcasting rates. In specific, we first introduce acircular recording queue to online perceive time slot occupyingstatus. We then design a distributed time slot sharing (DTSS)algorithm and random index first fit (RIFF) algorithm, toefficiently share the time slot and make the on-line vehicle-slot matching, respectively. We prove theoretically the efficacy ofDTSS algorithm, and evaluate the efficiency of RIFF algorithmby using Matlab simulations. Finally, we conduct extensivesimulations considering various driving scenarios and resourceconditions to demonstrate the SS-MAC performance.

Index Terms—Vehicular ad hoc networks; medium accesscontrol; time slot sharing; resource management.

I. INTRODUCTION

DRIVING safety has been the number one priority inIntelligent Transportation Systems (ITS). Transmitting

warnings about dangerous driving conditions among vehiclesthrough vehicular ad hoc networks (VANETs) has emerged asa promising solution to enhance driving safety [1]. To enhancetraffic safety and efficiency in VANETs, vehicles and road-side units (RSUs) need to periodically broadcast Basic SafetyMessages (BSMs) [2] to all neighbors within one-hop. Bybeing constantly aware of the events in their surrounding en-vironment, high-priority safety applications such as pre-crashsensing, blind spot warning, emergency electronic brake lightsand so on [3], can be supported. However, the communicationchannel may witness excessive network load generated byhigh-frequency periodical broadcasts, especially under high-density situations. Without efficient control, the aggregate ofthese broadcasts will congest the channel, impairing reception

Copyright (c) 2015 IEEE. Personal use of this material is permitted.However, permission to use this material for any other purposes must beobtained from the IEEE by sending a request to [email protected].

F. Lyu, H. Zhu (Corresponding author) and M. Li are with the De-partment of Computer Science and Engineering, Shanghai Jiao Tong Uni-versity, Shanghai, China, 200240, (e-mail: [email protected], {hongzi, li-ml}@cs.sjtu.edu.cn).

W. Xu and X. Shen are with the Department of Electrical and ComputerEngineering, University of Waterloo, 200 University Avenue West, Waterloo,Ontario, Canada, N2L 3G1, (e-mail: {w74xu, sshen}@uwaterloo.ca).

H. Zhou is with the School of Electronic Science and Engineering, NanjingUniversity, Nanjing, China, 210093, (e-mail: [email protected]).

N. Zhang is with the Department of Computing Sciences, Texas A&MUniversity at Corpus Christi, 6300 Ocean Dr., Corpus Christi, TX. 78412,(e-mail: [email protected]).

performance and safety benefit. In addition, safety messagesare delay-sensitive and require ultra-low latency, however, it ishard to find the centralized infrastructure for the coordinationsof safety messages distribution in VANETs. To devise anefficient media access control (MAC) protocol for safetyapplications with reliability and minimum delay under thisstrict distributed and high dynamic networks, is critical andchallenging.

In the literature, various MAC protocols have been proposedfor VANETs. In particular, the EEE 802.11p [4] standard asan amendment to the existing IEEE 802.11a-2007 or Wi-Fi [5]standard, has been dedicated by the Federal CommunicationsCommission (FFC) as the MAC layer for vehicle-to-vehicle(V2V) and vehicle-to-infrastructure (V2I) communications inthe United States. Even though the IEEE 802.11 is widelyimplemented, but it dose not provide an efficient broadcastservice in VANETs due to the following reasons. First, thebasic MAC method of IEEE 802.11p is the same as the dis-tributed coordination function (DCF) of IEEE 802.11, whichbases on the carrier sense multiple access/collision avoidance(CSMA/CA) mechanism. In the CSMA/CA, when a nodewants to access the medium, it has to sense the channelfirst; if the channel is idle, the node can access the medium;otherwise the node has to perform random back-off. This kindof contention-based MAC may result in possible unboundeddelays [4]–[7], which cannot satisfy the real-time requirementof safety applications in VANETs. Second, in broadcast modeof 802.11p protocol, request to send (RTS) /clear to send(CTS) /acknowledgement (ACK) packets are removed to fa-cilitate real-time response, which leaves the hidden terminalproblem unsolved [8]. Beyond that, the delay and reliabilityperformance of 802.11p DCF-based beaconing have beenanalyzed in many studies, and the results demonstrate thatthe 802.11p MAC has serious issues with unbounded delayand channel congestion under high-density scenarios [9], [10].Specifically, the authors in the work [11] discover that eventhough the number of collisions is reduced by dynamicallyadjusting the contention window in 802.11p protocol, thepacket reception probability still never reaches 90% because ofthe randomness of CSMA-based schemes. In the observationsof the work [12], the normal delay of beacons may last 200 msand the value can be more than 500 ms and sometimes reaches1 s in a dense scenario. As a result, it can be concluded thatthe contention-based MAC 802.11p is unsuitable for real-timecommunications due to its nondeterministic features.

Recently, time division multiple access (TDMA) basedMACs have demonstrated their efficiency in VANETs [13]–



2

[16]. In TDMA-based MACs, time is partitioned into framesconsisting of a constant number of equal-length time slots andsynchronized among vehicle nodes. Each vehicle is guaranteedto access the channel at least once in each frame; the slottedchannel can guarantee the stringent time requirement of safetyapplications. Moreover, each vehicle not only transmits itsapplication data but also reports the status of time slots used byits one-hop neighbors; by doing so, vehicles can perceive up-to-date knowledge of two-hop neighbors and acquire a distincttime slot with them, then vehicles can have the ability to detectcollisions and avoid the hidden/exposed terminal problemswithout RTS/CTS/ACK schemes. However, in existing TDMAMACs, a high constant beaconing rate, normally 10 Hz, isconfigured for all nodes, incurring serious scalability problemsfor resource allocation. Specifically, when the node density islow, scarce channel resources [17], [18] are wasted due tothe unnecessary broadcasting. According to the vehicle safetycommunications report of U.S. Department of Transportation[3], there are really distinct safety applications with the broad-casting rate ranging from 1 Hz to 10 Hz. Allocating excessresources to every vehicle for unnecessary broadcasting notonly wastes channel resources but also increases the possibilityto interfere others. On the other hand, due to the spatial reuseconstraint of TDMA 1, when the vehicle density is high, suchas at intersections, the slot starving problem may occur with aresult exposing unfairness of medium access among vehicles.For instance, when some vehicles at the intersection havefully occupied time slots in every frame, subsequent enteringvehicles may have no time slot to choose for transmission,then the time slot acquisition failures will last for a longtime, resulting in a huge medium access delay. To make thingsworse, if vehicles with low-priority 2 safety requirements havesuccessfully occupied time slots while vehicles with high-priority safety requirements have no chance for transmission,this extreme unfairness of medium access will badly impair thedriving safety. Just like traffic management in real life, whenthere is a traffic jam, the vehicles with more importance, e.g,police cars or ambulances, have a higher priority of passingthrough; the medium access control should also support thiskind of fairness when the channel is saturated. Moreover, todesign dynamic beaconing approaches for safety messages,beacon rate control is the main beaconing category [19]–[21]in VANETs; all these application layer approaches need ascalable and flexible MAC protocol [22], [23] to support.

For the aforementioned considerations, we propose SS-MAC, a novel time slot-sharing MAC for safety messagesbroadcasting in VANETs, with reliability and minimum delay.In SS-MAC, broadcast requirements of safety applications areperiodic with different rates, it is profitable to make multiplevehicles alternately broadcast on a same time slot throughinerratic coordination. To capture the periodic characteristicsof safety applications over time slots, we first introduce acircular recording queue to online perceive time slots occu-pying status. The circular recording queue records the time

1For reliable transmission without collisions, vehicles in two-hop commu-nication ranges should not use the same time slot.

2The priority is judged by the broadcast rate requirement in this paper.

slot status (occupied or vacant) during the latest K successiveframes; a suitable K recording queue can help perceive theseasonal occupied behaviors on each time slot. Based on theseinformation, we design a distributed time slot sharing (DTSS)approach, to decide whether the time slot can support thesharing for a certain periodical broadcasting requirement, andhow to share a time slot in an efficient way. Specifically, wepresent the precondition of a time slot sharing among vehicles;to satisfy the precondition, we propose normalizing cycles ofvehicles for consolidated sharing; we then define feasibilityparameter and sharing potential parameter as elaborated inSection III to motivate DTSS design with perfect sharingproperty. After that, we design a random index first fit (RIFF)algorithm based on the heuristic packing method, to makeon-line vehicle-slot matching with maximizing the resourceutilization of the network. We prove theoretically the efficacyof DTSS algorithm design and evaluate the efficiency ofRIFF algorithm by using Matlab simulations. In addition,we conduct extensive simulations considering various drivingscenarios and resource conditions to demonstrate the overallSS-MAC design in terms of delay ratios of the whole networkand each vehicle. In particular, the main contributions of thispaper are summarized as follows.• We design a novel time slot-sharing MAC, referred to as

SS-MAC, for safety messages broadcasting in VANETs.As SS-MAC supports the diverse periodical broadcastingrates, vehicles can access the medium according to theirsafety need. Hence, SS-MAC can work with a strongscalability in terms of channel resource managementunder all sorts of driving scenarios.

• To achieve the common agreement of the time slot shar-ing, we devise a distributed time slot sharing approach,called DTSS, to regulate the time slot sharing processamong vehicles. It can maximize the sharing potential ofeach certain time slot.

• We develop the RIFF algorithm to assist the vehicleselecting a suitable time slot for sharing. It can not onlysatisfy the broadcasting requirement of every vehicle butalso optimize the resource utilization of the network.

The remainder of this paper is organized as follows. Wepresent the system model and TDMA MAC basics in SectionII. We propose SS-MAC protocol in Section III. We evaluatethe efficiency of the RIFF algorithm and conduct extensivesimulations to evaluate the performance of SS-MAC in SectionIV. Finally, we conclude this paper and discuss future workin Section V.

II. SYSTEM MODEL AND TDMA MAC BASICS

A. System Model

We consider the VANET scenario as shown in Figure 1,where nodes communicate with each other through wirelesscommunication. In the scenario, communication nodes includevehicles and RSUs, both of which need to broadcast informa-tion periodically for driving safety. As the RSUs can accessthe channel via the same MAC protocol like the vehicles, forconvenience, we call the node or the vehicle on behalf of theset of the RSU and the vehicle in this paper.



3

Wireless communications RSUsVehicles

THS

OHS

A B

C D

E

Frame

AB C D E

Frame

AB C D E

≈

Fig. 1: Illustration of the system model.

Wireless communications. Dedicated Short Range Com-munications (DSRC) module is used by each node for wire-less communication. The DSRC radio supports one ControlChannel (CCH) and multiple Service Channels (SCHs) withtwo optional bandwidths of 10 MHz and 20 MHz [24], [25].The CCH is used to transmit control information or high-priority short messages (such as periodic or event drivensafety messages) while the SCHs are used to transmit non-safety user application data. In general, the radio switchesthe channel between the CCH and the SCH; when in CCHIntervals (CCHI), all nodes tune to the CCH to send/receivehigh priority safety messages or to negotiate the followingusage of SCHs among nodes; when in SCH Intervals (SCHI),the nodes tune to the specific SCH for services accordingto the negotiation result [26]. If there are two antennas, thefirst one is tuned to the CCH and the second one is tunedto the SCH; safety applications and negotiation informationof SCHs usage are transmitted over the CCH. In both cases,the transmission over the CCH is the foundation of safetyapplications and multi-channel operations. For this purpose,we focus on the design of an efficient medium access controlon CCH for safety applications in this paper. To controlthe medium access on the CCH, the channel is set to aslotted/framed structure. Specifically, time is partitioned intoframes and each frame contains N number of slots; each timeslot is set to a equal-length duration for data transmission;before the data transmission, a node has to apply a vacanttime slot first. As shown in Figure 1, the elements in lightcolor denote vacant time slots while the elements in dark colordenote occupied time slots by vehicles. In addition, the channelis thought to be symmetric which has been evaluated in thework [27]. Thus, if the node x is in the communication rangeof the node y iff the y is in the communication range of x.Each radio in the network has the identical communicationcapability and the same communication range R.

Vehicles. Each vehicle in the network is equipped withat least one DSRC radio and a global positioning system(GPS) receiver. The GPS not only provides the locationinformation but also does the synchronization among vehicles.To synchronize vehicles, the pulse per second (1 PPS) signal

provided by any GPS receiver can be utilized. Specifically,the rising edge of the 1 PPS signal is aligned with the startof every GPS second with an accuracy within 100 ns; thisaccurate 1 PPS signal can be used as a common time referenceamong all the nodes. As a result, at any instant, each nodecan determine the index of the current slot within a frame[15], [28]. In addition, the stability of the GPS receivers localoscillator at each node can still support synchronization withaccuracy for a short time, coping with the temporary loss ofGPS signal. For driving safety, each vehicle has to broadcastinformation periodically with a different frequency based onthe safety application type and all the safety applications havea low tolerance of delay. According to the vehicle safetycommunications report of U.S. Department of Transportation,the frequency of safety applications can range from 1 Hz to10 Hz [3]. Each vehicle is identified by a MAC address aswell as a randomly generated short identifier (ID).

B. TDMA MAC Basics

Before transmission, each node needs to acquire a uniquetime slot and once a time slot is assigned, the node canuse it in all subsequent frames. As shown in Figure 1, allthe neighboring vehicles in the communication range of thevehicle A constitute the one-hop set (OHS) of the vehicle A.In addition, the two-hop set (THS) of the vehicle A refers toall the vehicles that can reach the vehicle A in two hops atmost. Before introducing our design, in this subsection, wesimply conclude some TDMA MAC basics which will alsobe adopted by SS-MAC.

Broadcasting additional frame information (FI). Dueto the lack of infrastructure in VANETs, there is generallyno centralized coordinator to do time slot allocation. Tocoordinate the time slot using in a distributed way, eachvehicle needs to collect (passively hear) and broadcast additioninformation, called the frame information, to its OHS so thateach vehicle can perceive the time slot occupying status. Theframe information transmitted by a vehicle includes the vehicleIDs and the corresponding time slots in its OHS. In doingso, each vehicle can know all the occupied time slots byits THS. Adding such extra coordination data in broadcastpackets is acceptable due to that, the total packet size of safetyapplications (normally 200-500 bytes [2]) is far less than thesize of MAC layer protocol data unit.

For a vehicle x, the following sets are defined:• Ncch(x): the set of IDs of its one-hop neighbors, which

are updated upon whether the node x has received packetsdirectly on the channel in the previous N slots. Inaddition, the node x needs to broadcast this informationwith application data during each transmission.

• N2cch(x): the set of IDs of its two-hop neighbors, indirect-

ly obtained from the packets transmitted by its one-hopneighbors, i.e.,

N2cch(x) = Ncch(x)

⋃{Ncch(y),∀y ∈ Ncch(x)}.

Time slot acquisition. Obviously, vehicles in the sameOHS should select different time slots to avoid transmissioncollisions. To eliminate the hidden terminal problem, vehiclesin the same THS should also choose distinct time slots even



4

they may not collide with each other directly. Specifically, inFigure 1, the vehicle A and C are in the same THS witha common one-hop neighboring vehicle B; if vehicle A andC transmit messages simultaneously, collisions will happenat the vehicle B. By collecting frame information from itsOHS neighbors, a vehicle can randomly choose a free timeslot from the idle time slot set. Under this scheme, once avehicle has the latest channel information of a frame time, thevehicle can make the slot acquisition decision. To maintainsuch information, for a running radio, it can update the channelinformation during each frame while a newly opened radio hasto listen to the channel for an entire frame first.

Detecting collisions. Once a vehicle successfully acquiresa time slot, it can use it in all subsequent frames. However,due to the lack of centralized controllers and moving charac-teristics of VANETs, transmission collisions encounter with arelatively high probability. There are two kinds of collisionscan be identified, i.e., access collision and merging collision[15]. The access collision happens when two vehicles in thesame THS try to acquire a same time slot then the collisionencounters at their common neighboring vehicles; while themerging collision happens when two vehicles which are intwo different THS originally, use the same time slot and moveinto a same THS finally due to the diverse mobility. To detectthe collisions, each vehicle checks the frame informationreceived from its OHS during the previous N − 1 time slots.Specifically, if packets received from all y ∈ Ncch(x) indicatethat x ∈ Ncch(y), it means that there is no other node in thetwo-hop ranges of x accessing the same time slot with thenode x; otherwise, the node x may collide with other duringthe last transmission. Once a collision is detected, the nodeshould release its original time slot and try to apply a newtime slot.

III. SS-MAC PROTOCOL DESIGN

A. Design Overview

To design a scalable slot-sharing TDMA MAC for diverseperiodical broadcasting rates under strict distributed and highdynamic vehicular networks, a high-precision coordinatingis in need. In the following subsections, we first introducea circular recording queue to perceive time slots occupyingstatus. We then design a distributed time slot sharing approach,called DTSS, to decide whether the time slot can supportthe sharing for a certain periodical broadcasting requirementand how to share a time slot in an efficient way. Afterthat, we design a RIFF algorithm based on the heuristicpacking method, to make on-line vehicle-slot matching withmaximizing the resource utilization of the network.

B. Perceiving Time Slots Occupying Status

Prior to share a time slot, vehicles need to perceive the timeslot using status. As safety applications periodically broadcastmessages under different rates, one possible solution is foreach vehicle to adopt a circular recording queue to recordthe most recently using status (one means occupied whilezero means free) of every slot in each frame. Specifically, foreach time slot i, i ∈ [1, N ], the vehicle initialises a circular

... i ... i

≈

... i

≈

... i ... i≈

... i

Time

Frame Frame Frame Frame Frame Frame

t 1 frames

n-1

t 1 frames

Fig. 2: Precondition of time slot sharing for two periodicalsafety applications.

recording queue, denoted as Qi = [qK−1, qK−2, ..., q0] whereqj for j ∈ [0,K − 1] is the (j + 1)th element in the queuecounted from right to left which means the using status ofslot i in the previous jth frame, and q0 means the using statusof the current frame. For each vehicle, at the end of eachframe, based on the updated time slots using status U(x) ofthe current frame, if the slot i is used by its THS neighborsin N(x), elements in Qi move one step towards left directionwhich means the value qj is replaced by qj−1 and the q0is set to the value one; otherwise, elements in Qi move onestep towards left direction and the q0 is set to the value zero.With the circular recording queue keeping the most recentlyK frame time slots occupying information, the sharing statusof every time slot can be perceived online and can be thecoordination guideline for vehicles to share time slots.

To design an appropriate queue size K, the followingtradeoff should be considered. On one hand, if the size ofthe queue is too small, the K frame time cannot cover acomplete broadcasting cycle of safety applications with a lowfrequency, which means that the circular recording queuehas no capability of collecting all the needed coordinationinformation. On the other hand, if K is too large, the pastK frame time used to obtain the slot occupying informationcan last long; however, during this long duration, the networktopology and the time slot sharing status may change due tothe solid high dynamic characteristics of VANETs. As a result,the outdated records may incur inaccurate decision of time slotsharing.

C. Distributed Time Slot Sharing Approach

In this subsection, we first show the precondition of a timeslot sharing among vehicles. To satisfy the precondition, wethen propose normalizing cycles of vehicles for consolidatedsharing. We finally elaborate the DTSS design, in whichthe feasibility parameter and sharing potential parameter aredefined to maximize the sharing potential of a time slot. Underthis design, we prove that DTSS can work with the perfectsharing property.

Precondition of time slot sharing. As broadcast require-ments of safety applications are periodic with different cycles,it is profitable to make multiple vehicles alternately broad-cast on a same time slot for efficient utilization of channelresources. For convenience, we name the safety applicationsas t-cycle applications in this work when the safety applicationhas a periodic broadcast requirement every t frame time (eachframe usually last 100ms).



5

Lemma 1: For t1-cycle and t2-cycle applications, 1 < t1 ≤t2 ≤ T 3, T is the maximum cycle value, if t2 = n ∗ t1,∀n = 1, 2, 3, ..., then these two applications can share a sametime slot.

Proof: As shown in Figure 2, the t1-cycle applicationbroadcast once at the slot i, i ∈ [1, N ], during every t1 frames;the deep color arrow is an integer pointer p1 pointing to theframe that the t1-cycle application will use the slot i in thisframe, p1 ∈ [0, t1 − 1]. For the t2-cycle application, it alsobroadcast on the slot i and the light color arrow is an integerpointer p2 pointing to the frame that the t2-cycle applicationwill use the slot i in this frame, p2 ∈ [0, t1−1] and p2 6= p1 atinitial stage 4. As t2 = n ∗ t1, after n∗ t1 frames, the t2-cycleapplication will broadcast a new message at the slot i and thevalue of p1 and p2 will not change. Without collisions underthis periodic activities, then the lemma is proved.

However, if t2 = n∗ t1+k, k ∈ [1, t1−1], then p2 could be(p2+k∗m)%t1, where % is the modulus operator. As a result,p1 has the possibility to collide with p2 or not, depending onthe value of t1, t2, p1 and p2. We do not consider this sharingsituations due to the lack of regularity.

Normalizing cycles for consolidated sharing. Lemma1 shows a precondition of slot sharing which also limitsthe sharing scopes and brings complexity for coordination,especially in distributed systems. Specifically, for periodicalapplications with various cycles, few of them can have multi-ple relationships in terms of cycles; moreover, for vehicles, itis hard to detect all broadcast cycles under a wide range. Onereasonable solution is to normalize the cycle of applicationsto a close value, which not only satisfies the time QoS ofsafety applications, but also would be convenient for sharinga time slot with other vehicles. To do it, we define a listof normalizing targets, denoted as N = [n1, n2, ..., nZ ],1 < n1 < n2 < ... < nZ ≤ T , where nz for z ∈ [1, Z]is the zth target in the list and nz+1 = X ∗ nz , X is aninteger and X > 1.

Rule 1: For a t-cycle safety application, ∀t ∈ [1, T ], thecycle of the application is normalized as follows

t =

1 t < n1;nz nz ≤ t < nz+1;nZ nZ ≤ t ≤ T.

(1)

After normalization, periodical applications can share a timeslot with each other once the time slot has enough capacity. Tocollect needed coordination information, the size K of circularrecording queue can be set to the value nZ . Figure 3 is anexample of time slot sharing, where Z = 3 and n1 = 2,n2 = 4, n3 = 8. In the figure, a 2-cycle application, a 4-cycleapplication and a 8-cycle application are sharing a time slot i,i ∈ [1, N ]. By checking the circular recording queue of sloti, vehicles can perceive the sharing status of the slot and findthat it can still support sharing for another 8-cycle application.

DTSS algorithm design. After normalizing the cycle ofapplications, the other issue is how to coordinate among

3As 1-cycle applications need to broadcast during each frame, they can notshare a time slot with other applications.

4When two periodical broadcast applications share a time slot, they shoulduse the time slot at different frame; otherwise collisions will happen.

Time

8 frames cycle

≈

i i i i i i i i i i i i i i i i

8 frames cycle2-cycle 4-cycle 8-cycle

Fig. 3: An example of time slot sharing, where Z = 3 andn1 = 2, n2 = 4, n3 = 8.

vehicles to share a time slot not only without collisions, butalso fully utilizing the time slot. As the size K of circularrecording queue is set to the biggest cycle nZ , the occupyingstatus of K frames for a slot will repeat in the followingsubsequent K frames and vehicles can coordinate to share atime slot just based on the information of the circular recordingqueue. To regulate the usage of a time slot, we design theDTSS algorithm.

Definition 1: (Item size) For a t-cycle safety application,the item size α of the application is α = 1

t , α ∈ (0, 1].Definition 2: (Slot capacity) For a specific time slot, the

capacity C of the time slot is calculated by the number ofidle elements in the circular recording queue to the total sizeK and the capacity of a free time slot is C = 1.

Remark 1: For a safety application with a size of α, if theapplication has a chance to share the time slot, iff the remaincapacity of the time slot is equal or bigger than the size α,i.e.,

C ≥ α. (2)

For a t-cycle application, after normalizing the cycle to avalue nz , we define a list of serial numbers for the recordingqueue under this cycle by modulus operator, i.e, for an elementqj in the queue, j ∈ [0,K−1], the serial number sj = j%nz ,sj ∈ [0, nz − 1], is defined.

Definition 3: (Feasibility parameter) We define the feasibil-ity parameter fzj for each element in the circular recordingqueue under different cycle values, j ∈ [0,K − 1] andz ∈ [1, Z]. For a given z and j, its serial number satisfiessj = j%nz , for ∀{x|x%nz = sj , x ∈ [0,K − 1]}, if allqx = 0, then fzj is set to the value one, otherwise it is set tothe value zero, i.e.,

fzj =

{1,∀{x|x%nz = sj , x ∈ [0,K − 1]}, qx = 0;0, elsewhere.

(3)

Remark 2: If a nz-cycle application can share a time slot,iff there is a j ∈ [0, nz − 1] and fzj = 1, i.e.,

∃j ∈ [0, nz − 1], s.t., fzj = 1. (4)

If the application chooses the jth element in the recordingqueue to share the time slot, then all the elements ∀{x|x%nz =sj , x ∈ [0,K − 1]} will by occupied by the application todemand its requirement. Following Remark 1 and Remark 2,vehicles can share a time slot without collisions. However,choosing an appropriate value j in Remark 2 determines thesharing efficient of the time slot. For instance, Figure 4 shows



6

q 7 q 4q 5 q 2q 3 q 0q 1

Occupied Idle

Circular recording queue

q 6

01 01 01 01

23 01 23 01

67 45 23 01

modulus operator on n1



Fig. 4: An example of a circular recording queue and thecorresponding modulus operators on different cycle values,where K = 8 and n1 = 2, n2 = 4, n3 = 8.

an example of the circular recording queue of a specific timeslot and the corresponding modulus operators on differentcycle values, where the size of recording queue K is set to be8 and n1 = 2, n2 = 4, n3 = 8 respectively. In addition, thedark block denotes the occupied status while the white blockdenotes the idle status and the remain capacity of the timeslot is calculated by the number of idle blocks to the total sizeK, i.e, 3

8 . According to the capacity restriction of Remark 1,only the 4-cycle or 8-cycle applications can be supported bythis time slot. For a 8-cycle application, the element in therecording queue with a serial number 0, 2, 4 can satisfy itsdemand; if the application chooses to occupy the serial number2, then the remain capacity of the time slot can still support a4-cycle application, otherwise the time slot can only support8-cycle applications if the number 0 or 4 is chosen.

Remark 3: For an jth element in the recording queue, iffzj = 0, then fz−1j = 0, z ∈ [2, Z].

Remark 3 indicates that, if the jth element can not supportsharing for the nz-cycle applications, then it has no potentialfor supporting applications with a higher frequency. Consid-ering this, we define the sharing potential parameter pj foreach element in the recording queue as follows.

Definition 4: (Sharing potential parameter) For an jthelement in the recording queue, the sharing potential parameterpj is the maximum item size of the nz-cycle applicationsthat the corresponding value of feasibility parameter satisfiesfzj = 1, i.e.,

pj =1

nz, z = min{∀z ∈ [1, Z], s.t., fzj = 1}. (5)

In addition, the sharing potential of the time slot p is themaximum pj , for j ∈ [0,K − 1], i.e.,

p = max{pj , j ∈ [0,K − 1]}. (6)

Remark 4: If a nz-cycle application can share a time slot,iff the sharing potential of the time slot p satisfies p ≥ 1

nz.

Rule 2: For a nz-cycle application, if exists a list J ofelements in the recording queue(j ∈ [0, nz − 1]) can satisfyits sharing demand according to the limit of Remark 1 andRemark 2, DTSS algorithm chooses the element with theminimum value of sharing potential parameter from the listJ .

As shown in Figure 4, for a 8-cycle application, the 0th, 2thand 4th element can satisfy its requirement with the sharing

potential parameter p0 = 14 , p2 = 1

8 , p4 = 14 respectively.

According to Rule 2, the 2th element will be chosen andthe application can access the channel at this time slot afterwaiting for (nz − 1 − j) frames with a repeating cycle of8 frames. DTSS algorithm for distributed efficient time slotsharing is shown in Algorithm 1.

Algorithm 1 DTSS algorithm for a time slot sharing.

Input: Qi and cycle tOutput: waiting frames w

1: Initialize: w = −1, pmin = 12: Normalize: t, t← nz3: if Ci ≥ αz then4: for j ∈ [0, nz) do5: if fzj == 1 then6: if pj ≤ pmin then7: pmin = pj8: w = nz − 1− j9: return w

Definition 5: (Perfect sharing) We define the perfect sharingproperty for a time slot. It has the following feature; for a timeslot with the capacity C and a application with the normalizedcycle nz(z ∈ [1, Z]), if C ≥ 1

nz, then the time slot can support

sharing for the application.Lemma 2: DTSS algorithm can guarantee each time slot

with the perfect sharing property 5.Proof: For a time slot, it can support sharing for n1

numbers of n1-cycle applications and a group elements fora nz-cycle application can be divided into X groups elementsfor X nz+1-cycle applications sharing, nz+1 = X ∗ nz .Considering a time slot is being shared by the nz-cycleapplication with the number of xz , xz ≥ 0, z ∈ [1, Z] and

Z∑z=1

xz1

nz≤ 1. (7)

According to Rule 2 choosing the minimum sharing potentialelement, it can guarantee that if the group elements for anz-cycle application still have the space for a nz+1 appli-cation, the nz+1 application will choose elements from theremaining and will not occupy elements which have poten-tials for other nz-cycle applications. Based on this rule, theelements occupied by X numbers of nz+1-cycle applicationsfinally can be combined to a group elements for a nz-cycleapplication. Then the occupying status of a time slot can berepresented by the ni-cycle application with the number of yi,0 ≤ yi < X(y1 < n1), i ∈ [1, Z]. Under this considering, thecapacity of the time slot C =

∑Zi=1 li

1ni

and the li satisfies

li =

n1 − 1− y1 i.= 1;

X − 1− yi 1 < i < Z;X − yi i = Z.

(8)

To support the requirement of a nj-cycle application, if ∃i ∈[1, j], s.t., li > 0, apparently the time slot can support sharingfor the application; if ∀i ∈ [1, j], li = 0, according to the

5The Lemma 2 demonstrates the efficacy of the DTSS algorithm design intheory.



7

expression of the capacity C and the limit of Eq. (8), thus theC < 1

njwhich violates the condition in Definition 5. Then the

lemma is proved.

D. On-Line Vehicle-Slot Matching Approach

Given the recording queue information of a time slot, DTSSalgorithm can determine whether the time slot can satisfy thetime QoS of applications and how to share a time slot greedilyto maximize future sharing potential. However, in practical,the medium is set to be numbers of time slots and how toselect a satisfied time slot for nodes has an impact on theresource utilization of the network. According to Remark 4,each time slot has a sharing potential value, and only the itemsize of the application is smaller than the potential value thenthe application can use the slot. Inspired by this, the on-linevehicle-slot matching problem can be modeled as a on-linebin packing problem, where each application has a item sizeand each free time slot has a full potential 1. The on-linebin packing problem is a well-known NP-hard problem; someclassical on-line algorithms are proposed such as WF (Worst-Fit), BF (Best-Fit), FF (First-Fit) and so on; these classicalalgorithms can be generalized to the Any-Fit and the AlmostAny-Fit classes [29].

Any-Fit constraint: If B1, B2, ......Bj are the currentnonempty bins, the current item will be packed into Bj+1

iff it does not fit in any of the bins B1, B2, ......Bj .Almost Any-Fit constraint: If B1, B2, ......Bj are the

current nonempty bins and the Bk(k ≤ j) is the unique binwith the smallest content, the current item will be packed intoBk iff it does not fit in any of the bins to the left of Bk.

The class of on-line heuristics that satisfies the Any-Fitconstraint will be denoted by AF and the class of on-linealgorithms satisfying both constraints above will be denotedby AAF .

Theorem 1 (Johnson [30]): For every algorithm A ∈ AF ,

R∞FF = R∞BF ≤ R∞A ≤ R∞WF6. (9)

Theorem 2 (Johnson [30]): For every algorithm A ∈ AAF ,

R∞A = R∞FF . (10)

Theorem 1 and 2 demonstrate that the FF and BF algorithmcan achieve the lowest worst-case ratio compared with algo-rithms in AF and AAF class. In FF algorithm, the currentitem will be packed into the first nonempty bin which it fits,if no such nonempty exists, the algorithm will open a new binto pack the item. While the BF algorithm packs the currentitem into an open bin of largest content in which it fits; ifno such nonempty exists, the algorithm will open a new bin.However, during time slot acquisitions, there are many nodesmay acquire time slots simultaneously; the fitting results fordifferent nodes of FF or BF have a high possibility to be thesame value, which may incur collisions when more than twonodes are assigned to a same time slot. To solve this, wedesign a new heuristic algorithm called RIFF based on FF. If

6R∞A is the asymptotic worst-case ratio (or asymptotic performance ratio,

APR) and the number is the value that packings produced by algorithm Acompare to optimal packings in the worst case .

U = {U1, U2, ......, Uj} is the current used time slot list andUi(i ∈ [1, j]) is the ith element in the list U , RIFF algorithmfirstly generates a random variable k(k ∈ [1, j]) to index theelements in the U as follows

I =

{i+ j − k + 1 i < k;i− k + 1 i ≥ k, (11)

where the value I is the index of the element. RIFF algorithmthen allocates the lowest indexed time slot which can satisfythe time requirement of the application to the current node;if no such used time slot exists, RIFF will randomly choosea free time slot for the node. Apparently, RIFF algorithmsatisfies the Any-Fit constraint and meets the demand of theAlmost Any-Fit constraint at the most of time. RIFF algorithmfor on-line vehicle-slot matching is shown in Algorithm 2.

Algorithm 2 RIFF algorithm for on-line vehicle-slot matching.

Input: U , S and cycle tOutput: time slot s and waiting frames w

1: Initialize: s = 0, w = −12: if t == 1 then3: s← random(S − U)4: w = 05: else6: k = random(1, len(U))7: for i ∈ (1, len(U)) do8: index = i9: if index < k then

10: index = index+ len(U)− k + 111: else12: index = index− k + 113: w = DTSS(QUindex

, t)14: if w 6= −1 then15: s = Uindex16: break17: if s == 0 then18: s← random(S − U)19: w = 020: return s,w

IV. PERFORMANCE EVALUATION

In this section, we evaluate the performance of RIFF algo-rithm by using Matlab and conduct simulations to evaluate theefficiency of SS-MAC design.

A. Evaluating RIFF Algorithm

As the efficiency of DTSS algorithm has been theoreticallyproofed in the subsection III-C, we only evaluate the proposedon-line vehicle-slot matching approach. We compare RIFFalgorithm with the following two alternative approaches.• Random Fit Approach. In this approach, each node

randomly chooses a time slot which can meet its demand.• FF Approach. FF approach allocates the current node

with the lowest nonempty indexed time slot which cansatisfy its sharing requirement, assuming there is such



8

TABLE I: Parameters used to evaluate the RIFF algorithm.

Parameters ValueNumber of time slots [1, 100]

Cycles of applications (in frames) [1, 10]Number of vehicles [1, 1000]

Matching rounds [1, 500]Simulation times 50

0 30 60 90 120 150 180 210 240 270 300 330 360 3900

10

20

30

40

50

60

70

80

90

100

The number of matched vehicles

The

num

ber o

f occ

upie

d tim

e sl

ots

Random FitTheoreticalFFRIFF

Fig. 5: The comparison of online matching results by differentalgorithms.

a time slot; if no such a time slot exists, FF approachassigns a free time slot to the current node.

Random Fit approach is easily implemented while FF ap-proach has been proofed that it can achieve the best perfor-mance in terms of worst-case properties among all on-linematching heuristics. After the matching process, all nodesadopt DTSS algorithm to share a specific time slot in allmatching approaches. The detailed simulation parameters areshown in Table I. Each frame is set with 100 time slots inthe environment; the application cycle t of each vehicle israndomly generated ranging from 1 to 10, where the unit isa frame duration time; matching algorithms try to assign thetime slot for a vehicle at each round. We consider the followingtwo metrics to evaluate the proposed matching approaches:

1) Number of occupied time slots. It means the neededtime slot resources to guarantee the QoS of matchedvehicles.

2) Sharing efficiency. It refers to the ratio of the sharingcapacity to the total capacity of a time slot.

Figure 5 shows the time slot occupied results under differentmatching algorithms; a theoretical value 7 is added for a bettercomparison. We can have the following three observations.First, RIFF algorithm can achieve the similar performancewith FF algorithm as their two curves are tightly closed.Second, when resources are sufficient, RIFF algorithm canoutperform the Random Fit algorithm; for instance, when thereare 100 vehicles in the environment, Random Fit algorithmneeds about 66 time slots for data transmission while thenumber of needed time slots can be reduced to only 30 byadopting RIFF algorithm and the value is very close to thetheoretical value 26. More than half of resources are saved

7The theoretical value is just the sum size of matched vehicles. Apparently,this value can not be achieved in practical and is just a comparing reference.

0 20 40 60 80 100 120 140 160 180 2000.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

The number of matched vehicles

The

ave

rage

sha

ring

effic

ienc

y of

tim

e sl

ots

Random FitRIFF

The sharing efficiency gap during the matching process

Fig. 6: The sharing efficiency comparison between the RIFFand the Random Fit algorithm.

which greatly improves the utilization of resources. Third,when meets the resources shortage condition, RIFF algorithmcan serve much more vehicles comparing with Random Fitalgorithm; in specific, RIFF algorithm can support about 320vehicles with 100 time slots while Random Fit algorithmcan only support about 240 vehicles under the same resourcecondition.

Figure 6 shows the average sharing efficiency of time slotsduring the matching process. As the curve of FF algorithmis tightly close to RIFF, we just show the curve of RIFFfor clear observations. We can have the following two mainobservations. First, the average sharing efficiency is alwayshigher under RIFF algorithm compared with the values underRandom Fit algorithm. Second, during the matching process,most sharing efficiency values achieved by RIFF algorithmare above 98% while most values achieved by Random Fitalgorithm are under the 50%; this is also the main reasonthat RIFF can greatly improve the utilization of resourcescomparing with Random Fit algorithm.

B. Evaluating the SS-MAC Design

In this subsection, we conduct extensive simulations toevaluate the efficiency of SS-MAC, considering various roadscenarios and resource conditions.

Simulation Setup. We use the Simulation of Urban Mo-bility (SUMO) [31] to conduct simulations evaluating theperformance of SS-MAC. We consider two typical VANETroad scenarios, i.e., highways and urban surface roads. In thehighway scenario, a bidirectional 8-lane highway with 10 kmlong is set; each direction contains four lanes with a speedlimit of 60 km/h, 80 km/h, 100 km/h and 120 km/h,respectively. In the urban scenario, four bidirectional 6-laneroads with 4 km long converge at an intersection; each ofthe three lanes in one direction is given a speed limit of50 km/h, 60 km/h and 70 km/h respectively; traffic lightsare set at each inbound road segment at the intersection withthe duration of green light being 20 s.

In both scenarios, vehicles have different performance pa-rameters in terms of maximum velocity, acceleration abil-ity and deceleration ability. Ten different sets of vehicle



9

TABLE II: Attributes of vehicles in simulations.

Attributes Values DescriptionmaxSpeed [80, 240] The vehicle’s maximum velocity (in km/h).

accel [1.0, 5.0] The acceleration ability of vehicles (in m/s2).decel [3.0, 10.0] The deceleration ability of vehicles (in m/s2).length [4.0, 7.0] The vehicle’s length (in m).

minGap [3.0, 10.0] The minimum offset to the leading vehicle when standing in a jam(in m).car-following model Krauss The model used for car following.lane-changing model LC2013 The model used for changing lanes.

sigma [0.5, 1.0] The car-following model parameter defining the driver imperfection (between 0 and 1).impatience [0.5, 1.0] Willingness of drivers to impede vehicles with higher priority (between 0 and 1).

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Delay ratio during each second of the whole environment

CD

F

SS−MACAggressive MACConservative MAC

(a) With N = 40

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1


CD

F


(b) With N = 50

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1


CD

F


(c) With N = 60

Fig. 7: CDFs of delay ratio during each second of the whole environment under different resource conditions in the highwayscenario.

parameters are configured according to the main types ofvehicles on the market. Vehicles are driven under the Krausscar-following model and the LC2013 lane-changing model.In addition, driver imperfection and impatience parametersare also set to introduce the human factor into simulations.Table II summarizes the attributes of vehicles in simulations.

Vehicles are generated at the entrance of each road 8 withdifferent rates to mimic normal traffic conditions, i.e., high-way (10 vehicles/lane/minute), urban (6 vehicles/lane/minute).Each vehicle randomly chooses a performance parameterconfiguration and the destination road; when vehicles arriveat the destination, they will leave the simulation system. Inall simulations, the transmission range R is set to be 300 maccording to the observation that 802.11p-compatible onboardunits can support reliable data transmission within 300 m[25]. Following the vehicle safety communications report ofU.S. Department of Transportation [3], the time duration ofa frame is set to be 100 ms satisfying the highest frequencyrequirement of safety applications and the application cyclet (in frames) is set to be [1,10]. Table III summarizes thesimulation parameters 9.

Methodology. We compare the proposed slot sharing SS-MAC with the following two reasonable candidate MACschemes.

• Aggressive MAC scheme. In this scheme, once a vehiclehas acquired a time slot successfully, it will broadcastits information during each frame no matter what itsapplication cycle is. By doing so, vehicles within the

8The highway can be with 8 entrance lanes while the urban is 3∗4 entrancelanes.

9Note that, the numbers of running vehicles are recorded at saturated trafficconditions and the number is larger in urban scenario due to the traffic lights.

TABLE III: Simulation parameters.

Parameters Highway UrbanRoad length 10 km 4 km

Number of road segments 1 4Number of intersections 0 1

Number of lanes on each road 8 6Speed limit in lanes (in km/h) [60, 120] [50, 70]

Transmission range 300 m 300 mFrame duration 100 ms 100 ms

Cycles (in frames) [1, 10] [1, 10]Slot duration 1 ms 1 ms

Number of slots (40, 50, 60) (80, 90, 100)Loaded vehicles 1630 1360Running vehicles 400− 500 500− 600Simulation time 1000 s 1000 s

same THS can know its time slot acquisition and willnot try to occupy the same time slot.

• Conservative MAC scheme. On the contrary, in thisscheme, a vehicle just transmits its information when inneed according to its application cycle. The advantage ofthis scheme is that only necessary resources are used.

We consider the delay ratio as the metric to evaluate the per-formance of SS-MAC. It refers to the number of unsuccessfultransmissions 10 to the total number of needed transmissionsaccording to the cycle of applications.

Performance comparison. In this subsection, we compareSS-MAC with other alternative MAC schemes under highwayand urban scenarios. We focus on the delay ratio of the wholenetwork and the delay ratio of each vehicle to consider theperformance of SS-MAC.

10An unsuccessful transmission happens due to two reasons; one is themedium access collision during the transmission, the other one is the casewithout medium to access due to the unfairness of the resource allocation.



10

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Delay ratio of each vehicle

CD

F


(a) With N = 40

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1


CD

F


(b) With N = 50

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1


CD

F


(c) With N = 60

Fig. 8: CDFs of delay ratio of each vehicle under different resource conditions in the highway scenario.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1


CD

F


(a) With N = 80

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1


CD

F


(b) With N = 90

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1


CD

F


(c) With N = 100

Fig. 9: CDFs of delay ratio of each vehicle under different resource conditions in the urban scenario.

Figure 7 shows the cumulative distribution functions (CDFs)of delay ratio during each second of the whole network underdifferent resource conditions in the highway scenario. We havethe following three main observations. First, it can be seenthat SS-MAC outperforms other schemes under all resourceconditions; for instance, when N = 40, more than 80% delayratios are smaller than the value of 0.2 in SS-MAC, whilethe value can reach about 0.43 in Aggressive MAC schemeand 0.7 in Conservative MAC scheme. Second, with moreresources, although all MAC schemes can ease the delayratio, the bending nature of curves do not change; specifically,the curve of SS-MAC is convex where the fast growth rateof y locates at the smaller x parts while the curves of theother two MACs are concave where the fast growth rate ofy locates at the larger x parts. For instance, within the valueof 0.1, the proportion can increase to about 40%, 80% and90% respectively under different resource conditions in SS-MAC, while the proportion just increase to 2%, 15% and 40%respectively in Aggressive MAC scheme and 3%, 10% and30% respectively in Conservative MAC scheme. The nature ofconcavity and convexity demonstrates the inherent efficiencyof SS-MAC. Third, the CDF gaps between SS-MAC with theother two MACs decrease with the improving of the resourceconditions; it demonstrates the stability of the SS-MAC in faceof tense resource conditions.

Figure 8 shows CDFs of delay ratio of each vehicle underdifferent resource conditions in the highway scenario. Wehave the following two main observations. First, SS-MAC canachieve the lowest delay ratio under all resource conditions; forexample, when N = 40, more than 90% vehicles can achieve

the delay ratio smaller than the value of 0.3 in SS-MAC, whilethe value can be about 0.66 and 0.88 in Aggressive MACscheme and Conservative MAC scheme respectively. Second,the curves of SS-MAC are steep while the curves of the othertwo MACs are flat; specifically, to reach the proportion nearly100%, the delay ratios can range from the value 0 to the valueabout 0.68, 0.5 and 0.41 respectively under different resourceconditions in SS-MAC, while the values reach to 1, 1 and0.74 respectively in Aggressive MAC scheme and 0.97, 0.9and 0.72 respectively in Conservative MAC scheme; the curveof the flat shows the unfairness the medium access.

Similar results can also be observed in urban scenarios. wejust show CDFs of delay ratio of each vehicle in Figure 9due to the space limitations. From the figure, we can easilyobserve that SS-MAC can achieve supreme performance underall resource conditions.

V. CONCLUSIONS AND FUTURE WORK

In this paper, we have proposed a novel time slot-sharingMAC, named SS-MAC, to support diverse beacon rates forsafety applications in VANETs. In specific, we have firstintroduced a circular recording queue to perceive occupancystates of time slots online and then designed a distributedtime slot sharing approach called DTSS to efficiently share aspecific time slot. In addition, we have developed the randomindex first fit algorithm, i.e., RIFF, to assist vehicle selectinga suitable time slot for sharing with maximizing the resourceutilization of the network. We have theoretically proved theefficacy of DTSS design, and evaluated the efficiency ofRIFF algorithm by using Matlab simulations. Finally, we have



11

conducted extensive simulations considering various drivingscenarios and resource conditions to demonstrate SS-MACdesign; particularly, delay ratios of the overall system andeach vehicle can be greatly reduced in highway and urbanscenarios under all kinds of resource conditions. For our futurework, we will investigate the dynamical resource assignmentbased on safety-awareness for safety applications and non-safety applications.

ACKNOWLEDGMENT

This research was supported in part by National Nat-ural Science Foundation of China (Grants No. 61472255,61420106010, 61373157, 61471236, 91638204, 61772340)and Natural Sciences and Engineering Research Council ofCanada (NSERC).

REFERENCES

[1] N. Lu, N. Cheng, N. Zhang, X. Shen, and J. W. Mark, “ConnectedVehicles: Solutions and Challenges,” IEEE Internet of Things Journal,vol. 1, no. 4, pp. 289–299, Aug. 2014.

[2] SAE International, “Dedicated Short Range Communications (DSRC)Message Set Dictionary,” Nov. 2009.

[3] CAMP Vehicle Safety Communications Consortium and others, “Vehiclesafety communications project: Task 3 final report: identify intelligentvehicle safety applications enabled by DSRC,” National Highway TrafficSafety Administration, US Department of Transportation, WashingtonDC, Mar. 2005.

[4] IEEE Std 802.11p-2010, “Standard for Information Technology-Telecommunications and Information Exchange between Systems-Localand Metropolitan Area Networks-Specific Requirements Part 11: Wire-less LAN Medium Access Control (MAC) and Physical Layer (PHY)Specifications Amendment 6: Wireless Access in Vehicular Environ-ments,” pp. 1–51, Jul. 2010.

[5] IEEE Std 802.11-2007 (Revision of IEEE Std. 802.11-1999), “Stan-dard for Information Technology-Telecommunications and InformationExchange between Systems-Local and Metropolitan Area Networks-Specific Requirements - Part 11: Wireless LAN Medium Access Control(MAC) and Physical Layer (PHY) Specifications,” pp. 1–1184, Jue.2007.

[6] S. Subramanian, M. Werner, S. Liu, J. Jose, R. Lupoaie, and X. Wu,“Congestion Control for Vehicular Safety: Synchronous and Asyn-chronous MAC Algorithms,” in Proceedings of the Ninth ACM Inter-national Workshop on Vehicular Inter-networking, Systems, and Appli-cations. New York, NY, USA: ACM, Jun. 2012, pp. 63–72.

[7] Q. Ye, W. Zhuang, L. Li, and P. Vigneron, “Traffic-Load-AdaptiveMedium Access Control for Fully Connected Mobile Ad Hoc Networks,”IEEE Transactions on Vehicular Technology, vol. 65, no. 11, pp. 9358–9371, Nov. 2016.

[8] Q. Ye and W. Zhuang, “Token-Based Adaptive MAC for a Two-HopInternet-of-Things Enabled MANET,” IEEE Internet of Things Journal,vol. 4, no. 5, pp. 1739–1753, Oct. 2017.

[9] Q. Chen, D. Jiang, and L. Delgrossi, “IEEE 1609.4 DSRC multi-channeloperations and its implications on vehicle safety communications,” in2009 IEEE Vehicular Networking Conference (VNC), Oct. 2009, pp. 1–8.

[10] M. I. Hassan, H. L. Vu, and T. Sakurai, “Performance Analysis ofthe IEEE 802.11 MAC Protocol for DSRC Safety Applications,” IEEETransactions on Vehicular Technology, vol. 60, no. 8, pp. 3882–3896,Oct. 2011.

[11] Y. Mertens, M. Wellens, and P. Mahonen, “Simulation-Based Perfor-mance Evaluation of Enhanced Broadcast Schemes for IEEE 802.11-Based Vehicular Networks,” in VTC Spring 2008 - IEEE VehicularTechnology Conference, May. 2008, pp. 3042–3046.

[12] R. Reinders, M. van Eenennaam, G. Karagiannis, and G. Heijenk,“Contention window analysis for beaconing in vanets,” in 2011 7th Inter-national Wireless Communications and Mobile Computing Conference,Jul. 2011, pp. 1481–1487.

[13] M. Hadded, P. Muhlethaler, A. Laouiti, R. Zagrouba, and L. A. Saidane,“TDMA-Based MAC Protocols for Vehicular Ad Hoc Networks: ASurvey, Qualitative Analysis, and Open Research Issues,” IEEE Com-munications Surveys Tutorials, vol. 17, no. 4, pp. 2461–2492, Jun. 2015.

[14] F. Borgonovo, A. Capone, M. Cesana, and L. Fratta, “ADHOC MAC:New MAC Architecture for Ad Hoc Networks Providing Efficient andReliable Point-to-point and Broadcast Services,” Wireless Networks,vol. 10, no. 4, pp. 359–366, Jul. 2004.

[15] H. A. Omar and W. Zhuang and L. Li, “VeMAC: A TDMA-BasedMAC Protocol for Reliable Broadcast in VANETs,” IEEE Transactionson Mobile Computing, vol. 12, no. 9, pp. 1724–1736, Jun. 2013.

[16] X. Jiang and D. H. C. Du, “PTMAC: A Prediction-Based TDMA MACProtocol for Reducing Packet Collisions in VANET,” IEEE Transactionson Vehicular Technology, vol. 65, no. 11, pp. 9209–9223, Nov. 2016.

[17] H. Zhou, N. Cheng, Q. Yu, X. Shen, D. Shan, and F. Bai, “Toward Multi-Radio Vehicular Data Piping for Dynamic DSRC/TVWS SpectrumSharing,” IEEE Journal on Selected Areas in Communications, vol. 34,no. 10, pp. 2575–2588, Oct. 2016.

[18] H. Zhou, B. Liu, T. H. Luan, F. Hou, L. Gui, Y. Li, Q. Yu, and X. Shen,“ChainCluster: Engineering a Cooperative Content Distribution Frame-work for Highway Vehicular Communications,” IEEE Transactions onIntelligent Transportation Systems, vol. 15, no. 6, pp. 2644–2657, Dec.2014.

[19] S. A. A. Shah, E. Ahmed, F. Xia, A. Karim, M. Shiraz, and R. M. Noor,“Adaptive Beaconing Approaches for Vehicular Ad Hoc Networks: ASurvey,” IEEE Systems Journal, vol. PP, no. 99, pp. 1–15, Jul. 2016.

[20] N. Wisitpongphan, O. K. Tonguz, J. S. Parikh, P. Mudalige, F. Bai, andV. Sadekar, “Broadcast storm mitigation techniques in vehicular ad hocnetworks,” IEEE Wireless Communications, vol. 14, no. 6, pp. 84–94,Dec. 2007.

[21] Y.-C. Tseng, S.-Y. Ni, Y.-S. Chen, and J.-P. Sheu, “The Broadcast StormProblem in a Mobile Ad Hoc Network,” Wireless Networks., vol. 8, no.2/3, pp. 153–167, Mar. 2002.

[22] Z. M. Fadlullah, D. Takaishi, H. Nishiyama, N. Kato, and R. Miura, “Adynamic trajectory control algorithm for improving the communicationthroughput and delay in uav-aided networks,” IEEE Network, vol. 30,no. 1, pp. 100–105, Jan. 2016.

[23] Z. M. Fadlullah, Y. Kawamoto, H. Nishiyama, N. Kato, N. Egashira,K. Yano, and T. Kumagai, “Multi-hop wireless transmission in multi-band wlan systems: Proposal and future perspective,” IEEE WirelessCommunications, vol. PP, no. 99, pp. 1–6, Oct. 2017.

[24] K. Abboud, H. A. Omar, and W. Zhuang, “Interworking of DSRC andCellular Network Technologies for V2X Communications: A Survey,”IEEE Transactions on Vehicular Technology, vol. 65, no. 12, pp. 9457–9470, Dec. 2016.

[25] F. Lv, H. Zhu, H. Xue, Y. Zhu, S. Chang, M. Dong, and M. Li,“An Empirical Study on Urban IEEE 802.11p Vehicle-to-Vehicle Com-munication,” 13th Annual IEEE International Conference on Sensing,Communication, and Networking (SECON), pp. 1–9, Jun. 2016.

[26] “IEEE Draft Standard for Wireless Access in Vehicular Environments(WAVE) - Multi-Channel Operation,” IEEE Unapproved Draft StdP1609.4/D6.0, Mar. 2010.

[27] F. Bai, D. D. Stancil, and H. Krishnan, “Toward Understanding Char-acteristics of Dedicated Short Range Communications (DSRC) froma Perspective of Vehicular Network Engineers,” Proceedings of ACMMobiCom, Sep. 2010.

[28] R. Ben-El-Kezadri and G. Pau, “Timeremap: stable and accurate time invehicular networks,” IEEE Communications Magazine, vol. 48, no. 12,pp. 52–57, Dec. 2010.

[29] E. G. Coffman Jr, J. Csirik, G. Galambos, S. Martello, and D. Vigo,“Bin packing approximation algorithms: survey and classification,” pp.455–531, Jun. 2013.

[30] D. S. Johnson, “Fast algorithms for bin packing,” Journal of Computerand System Sciences, vol. 8, no. 3, pp. 272–314, Jun. 1974.

[31] DLR Institute of Transportation Systems, “Sumo: Simulation of urbanmobility,” http://www.dlr.de/ts/en/desktopdefault.aspx/tabid-1213/.



12

Feng Lyu received the BS degree in software fromCentral South University in 2013. He is pursuinghis Ph.D degree in the Department of ComputerScience and Engineering at Shanghai Jiao TongUniversity. From October 2016 to October 2017,he was a visiting Ph.D. student at the BroadbandCommunications Research (BBCR) group in theDepartment of Electrical and Computer Engineer-ing, University of Waterloo, Ontario, Canada. Hisresearch interests include vehicular ad hoc networksand cloud computing.

Hongzi Zhu received his Ph.D. degree from SJTUin 2009, and M.Eng. and B.S. from Jilin Universityin 2001 and 2004, respectively. He was a PostDoctoral Fellow at University of Waterloo and theHong Kong University of Science and Technology in2009 and 2010, respectively. He is now an associateprofessor at the Department of Computer Scienceand Engineering, Shanghai Jiao Tong University(SJTU). His research interests include wireless net-works, mobile sensing and mobile computing, andvehicular ad hoc networks. He received the Best

Paper Award from IEEE Globecom 2016. He is a member of IEEE ComputerSociety and IEEE Communications Society. For more information, please visithttp://lion.sjtu.edu.cn.

Haibo Zhou (M’14) received the Ph.D. degree ininformation and communication engineering fromShanghai Jiao Tong University, Shanghai, China, in2014. From 2014 to 2017, he worked as a Post-Doctoral Fellow with the Broadband Communica-tions Research Group, ECE Department,Universityof Waterloo. Currently, he is an Associate Professorwith the School of Electronic Science and Engi-neering, Nanjing University. His research interestsinclude resource management and protocol designin cognitive radio networks and vehicular networks.

Wenchao Xu received the B.E. and M.E. degreesfrom Zhejiang University, Hangzhou, China, in 2008and 2011, respectively. He is currently workingtoward the Ph.D. degree with the Department ofElectrical and Computer Engineering, University ofWaterloo, Waterloo, ON, Canada. In 2011, he joinedAlcatel Lucent Shanghai Bell Co. Ltd., where hewas a Software Engineer for telecom virtualization.His interests include wireless communications withemphasis on resource allocation, network modeling,and mobile data offloading.

Ning Zhang [M’15] ([email protected]) re-ceived his Ph.D degree from the University of Wa-terloo in 2015. He is now an assistant professorin the Department of Computing Science at TexasA&M University-Corpus Christi. Before that, hewas a postdoctoral research fellow at the Universityof Waterloo and at the University of Toronto. Heis the recipient of the Best Paper Award at IEEEGlobecom 2014 and IEEE WCSP 2015. He is alead guest editor of Wireless Communications andMobile Computing and International Journal of Dis-

tributed Sensor Networks, and a guest editor of Mobile Information System.His current research interests include next generation wireless networks,software defined networking, vehicular networks, and physical layer security.

Minglu Li received the PhD degree in computersoftware from Shanghai Jiao Tong University in1996. He is a full professor and the vice dean ofthe School of Electronics Information and ElectricalEngineering, the director of Network ComputingCenter at Shanghai Jiao Tong University. His re-search interests include grid computing, servicescomputing, and sensor networks. He is a memberof the IEEE.

Xuemin (Sherman) Shen (IEEE M’97-SM’02-F09) received the B.Sc.(1982) degree from DalianMaritime University (China) and the M.Sc. (1987)and Ph.D. degrees (1990) from Rutgers University,New Jersey (USA), all in electrical engineering.He is a University Professor and the AssociateChair for Graduate Studies, Department of Electricaland Computer Engineering, University of Waterloo,Canada. Dr. Shen’s research focuses on resourcemanagement, wireless network security, social net-works, smart grid, and vehicular ad hoc and sensor

networks. Dr. Shen served as the Technical Program Committee Chair/Co-Chair for IEEE Globecom’16, Infocom’14, IEEE VTC’10 Fall, and Globe-com’07, the Symposia Chair for IEEE ICC’10, the Tutorial Chair forIEEE VTC’11 Spring and IEEE ICC’08, the General Co-Chair for ACMMobihoc’15, Chinacom’07 and the Chair for IEEE Communications SocietyTechnical Committee on Wireless Communications. He also serves/served asthe Editor-in-Chief for IEEE Internet of Things Journal, IEEE Network, Peer-to-Peer Networking and Application, and IET Communications; a FoundingArea Editor for IEEE Transactions on Wireless Communications; an AssociateEditor for IEEE Transactions on Vehicular Technology, Computer Networks,and ACM/Wireless Networks, etc.; and the Guest Editor for IEEE JSAC,IEEE Wireless Communications, and IEEE Communications Magazine, etc.Dr. Shen received the Excellent Graduate Supervision Award in 2006, andthe Premiers Research Excellence Award (PREA) in 2003 from the Provinceof Ontario, Canada. He is a registered Professional Engineer of Ontario,Canada, an Engineering Institute of Canada Fellow, a Canadian Academy ofEngineering Fellow, a Royal Society of Canada Fellow, and a DistinguishedLecturer of IEEE Vehicular Technology Society and Communications Society.

Documents

SS-MAC: A Novel Time Slot-Sharing MAC for Safety Messages ... · Messages (BSMs) [2] to all neighbors within one-hop. By being constantly aware of the events in their surrounding